Would someone PLEASE lay out the bottom line of keyword research -- competing sites vs. page rank?

[thegoddessofeleven]

I've heard some debate on the forum about the proper way to do keyword research.

Of particular interest to me is keyword research and how it pertains to article marketing for the purpose of promoting affiliate products.

Although conventional advice suggests that an IMer should find long tail keywords and then search said keyword in quotes to get the number of competing sites and hope that that number is low, I've heard others say that the number of sites returned for a given keyword isn't relevant at all-- only the PageRank of the sites on the first page of search engine results for that keyword are really an issue. If your pagerank is higher than that of the sites returned for that keyphrase, you're golden.

I can understand the logic of the latter method. I decided to try to change my keyword research approach to incorporate it. Then it got confusing.

Some time ago, I published an article on EzineArticles targeting a particular keyphrase. With or without quotes, my article is the first thing returned by Google for that phrase.
http://www.prchecker.info ranks EzineArticles a 6/10. It ranks my article's URL a 0/10. The page listed for spots below my article in Google ranks an 8/10.

I'm now entirely confused. If PageRank is a reliable way to do keyword research, why am I placing higher for this phrase in Google than a site that outranks me?

Is there ANY benefit to evaluating the number of sites returned by an "exact phrase" search?

[Steve Diamond]

Your own experience with EZA shows clearly that the overall authority of the site (EZA) has some influence regardless of the PR of the specific page where your article appears. I'm reasonably sure that you couldn't have gotten the same result by placing the same article on a new site with no PR.

So one good rule of thumb for research is that if there are already pages on authority sites like EZA, Squidoo, Wordpress.com, etc., that rank well for your keyword, that makes it more likely that you can achieve similar results with well placed content on that type of site.

Other than that idea, I feel that the PR of competing sites is less important than other factors.

As for how to gauge the "real" number of competing sites, I prefer to Google allintitle:"keyword phrase" as opposed to either of the methods you mentioned. This generally returns only the sites that are specifically optimized to rank for that phrase.

You may want to have a look at my current WSO (see sig) for a more in-depth look at competitive SEO analysis.

I hope this helps.

Steve

[thegoddessofeleven]

Thank you for the input! I'll take a look at your WSO.

[newBum76]

I've been wondering about page rank myself lately...but all the guides I've been reading only seem to talk about monthly searches and picking keywords that have under 30,000 competing sites......so I assume those are the biggest two factors to worry about as far as keyword research. I was researching yesterday and stumbled upon a keyword with almost 10,000 monthly searches and only 290 competing sites for a very competetive niche.....I would imagine that one will do pretty well on its own without any backlinking, we'll see!

[Steven Carl Kelly]

Originally Posted by newBum76

but all the guides I've been reading only seem to talk about monthly searches and picking keywords that have under 30,000 competing sites......so I assume those are the biggest two factors to worry about as far as keyword research.

The guides you're reading are giving you flawed information about how to choose a niche and analyzing "competion". First, Google ranks pages and not sites. Second, the pages you need to "worry" about are all on the first page of Google. 10,000,000 pages or 10,000 pages... Google will only return a maximum of 1,000 results anyhow.

Concentrate on the QUALITY of the competition, not the quantity. I don't care if there are 100M pages competing for my keyword phrase: if all of the first ten or so pages stink, I have no problem trying to out-rank them. But if there are only 900 competing pages and every one of the top ten or so are mega authority sites, then I may opt not to wade into that fight.